containers/nri-plugins

off-the-shelf configuration results in warnings

Opened this issue · 0 comments

I'm trying to run topology-aware nri-plugin, and using the off-the-shelf configurations from this git. However, I found the following warnings in the topology-aware plugin log:

W0907 08:41:06.438974       1 resources.go:567] possible misconfiguration of reserved resources:
W0907 08:41:06.439054       1 resources.go:568]   root: allocatable <root allocatable: CPU: reserved:0 (allocatable: 51m), grantedReserved:949m, sharable:1-223 (allocatable:223000m), MemLimit: DRAM 1.97T>
W0907 08:41:06.439062       1 resources.go:569]   kube-system/nri-resource-policy-topology-aware-g5fxw/nri-resource-policy-topology-aware: needs 500 reserved, only 51 available
W0907 08:41:06.439067       1 resources.go:571]   falling back to using normal unreserved CPUs instead...

My environment is:

  • Ubuntu 22.04 with kernel 5.15.0-83-generic
  • containerd: v1.7.5
  • nri-plugin: git commit c587af4
  • K8S: v1.23.6, kubeadm setup, no special configuration, all out-of-box default configuration

May I ask what's the warning messages means? Does it matters?